Copyright 2007, The Johns Hopkins University and Lynda Burton. All rights reserved. Use of these materials permitted only in accordance with license rights granted. Materials provided “AS IS”; no representations or warranties provided. User assumes all responsibility for use, and all liability related thereto, and must independently review all materials for accuracy and efficacy. May contain materials owned by others. User is responsible for obtaining permissions for use from third parties as needed.

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike License. Your use of this material constitutes acceptance of that license and the conditions of use of materials on this site.

Application of Information Systems and Secondary Data

Lynda Burton, ScDJohns Hopkins University

Section A

Secondary Data

4

Application of Secondary Data Sources to HSRE

Pre-existing or pre-collected data is considered “secondary”as opposed to data collected specifically for a project which isconsidered “primary”There are a wide variety of secondary data sources

5

Secondary Data for HSRE: Several Major Categories

Archival data or record systemsOngoing “monitoring” systems/special repetitive “surveys”Management information systems (MIS)Billing transaction systems or other limited MISSpecial “one-shot” databases developed by other researchers

6

Advantages of Secondary Data

CheaperAvailable quicklyUsually available for long time periods, therefore useful in time series analysesMany are quite reliable and validOften can be used to target primary data collection

7

Disadvantages of Secondary Data

Available data are rarely perfect for your research needsReliability and validity problems may be presentRisk of breach of confidentiality exists

8

Archival Data Systems and Record Systems

Major goal is to provide an archival historical documentation of health-related occurrences or activitiesResponsibilities of those collecting the data usually stop afterthe database has been developedThese systems often have legal implications or satisfy some reporting requirementsMany problems exist with regard to uniformity

9

Examples of Archival Data Systems

Vital records (birth, death)HSCRC hospital discharge filesState-mandated “incident” reportsMedical records

10

Use of Medical Records in Research

Patient/client treatment records are the most ubiquitous data system in health care organizationsProvides the most detailed account of treatment process and patient clinical and demographic characteristics

11

Disadvantages of Use of Medical Records in Research

Record quality is inconsistentOften impossible to readInformation not uniformly reportedDifficult to retrieve compared to automated records

12

Management Information Systems

DefinitionA systematic method used to collect, process, store, retrieve, and transmit selected data on patients/clients, clinical activities, and financial transactions that provides some organizational personnel with information required to carry out a specific management function

13

Management Information Systems

Usually conceived with a specific organization or program−

Hence, mostly applicable to HSE

There are as many non-functional MIS’s as there are functional onesAn MIS is not necessarily automatedUnless information is used consistently, validity and reliability issues almost always existMIS data are usually timely (often “real time”)If evaluation is objective from beginning, HSE can theoretically be fully integrated into MIS

14

Examples of MIS

Total hospital information systemEncounter reporting systemHospital financial reporting systemsSome medical record systems−

Nursing home minimum data set

15

Billing/Transaction Data Systems

Limited for the most part to financial data (very little clinically significant data)Often only data available in machine-readable form on majority of health care transactionsCan be used for large scale analysesFairly reliable, both payor and payee are interested in accurate dataOften difficult to gain access

16

Examples of Billing Data Systems

Blue Cross recordsMedicare files−

National Claims History files

−

5% filesMedicaid filesHospital billing systems

Section B

National Surveys of Importance to HSR

18

Features of National Surveys

Sample drawn to be representative of large group (usually national population)Questions validatedData collection methods highly reliableOften panel survey, though some cohort surveysWith panel surveys, replacements added

19

National Health Interview SurveyNational Hospital Discharge SurveyNational Ambulatory Medical Care SurveyNHANESMedical Expenditure Panel SurveyMedicare Current Beneficiary Survey

Secondary Data Sources for HSRE: On-Going National Surveys

20

National Health Interview Survey

Principal source of information on health for U.S. populationContinuing survey with special studies addedHousehold interviewsNon-institutionalized, civilian populationProbability sample representative of the target population

21

National Health Interview Survey: Purpose

Provide national data on . . .−

Incidence of acute illness and accidental injuries

−

Prevalence of chronic conditions and impairments−

Extent of disability

−

Utilization of health care services−

Other health-related topics

22

0

1

2

3

4

5

6

7

All <15 15-24 25-44 45-64 65-74 75+Patient's Age

Off

ice

Vis

its

Per

Yea

r MaleFemale

Office Visits Per Year By Patient’s Age

Notes Available

23

*Overweight = BMI of 25 or more Percent of adults (standard error)

Selected BothCharacteristic Sexes Male Female

Age18-24 years 37.5 (0.72) 42.5 (1.06) 32.3(0.95)25-44 53.7(0.36) 64.3(0.48) 43.2(0.47)45-64 63.5(0.43) 71.5(0.61) 43.2(0.47)65-74 61.1(0.67) 67.0(0.93) 56.2(0.91)75 years and over 47.2(0.75) 50.0(1.31) 45.3(0.88)

EducationLess than high school graduate .....… 60.4(0.52) 62.0(0.75) 58.7(0.75)High school graduate ...……… 56.8(0.44) 64.7(0.65) 49.7(0.55)Bachelor of Arts,Science degree .....… 47.5(0.60) 59.0(0.89) 35.7(0.85)Master ’s,doctorate,medical degree ..... 42.4(0.99) 53.7(1.73) 29.2(0.96)

Poverty status

Below poverty level .......................... 56.9(0.66) 57.0(1.10) 56.6(0.83)1.00 –1.99 times poverty level .................. 58.7(0.60) 62.4(0.84) 55.4(0.78)2.00 –3.99 times poverty level ................... 56.7(0.43) 63.5(0.65) 49.7(0.58)

Notes Available

Percentage of Adults Who Were Overweight* by Selected Characteristics: U.S., 1997 –1998

24

0

5

10

15

20

25

30

35

40

< $5,000 $5000-$9,999

$10,000-$19,999

$20,000-$34,999

$35,000-$49,999

>$50,000

Family Income

Per

cen

t

All Ages

Under 65

Family Income by Age

Notes Available

25

National Hospital Discharge Survey

Data abstracted from medical records of discharges from short-stay, non-federal hospitalsCoded using ICD9-CM categoriesNumber, rate, and average length of stay, by age, geographic region, and genderNumber of discharges by first listed DX

26

Selected BothCharacteristic Sexes Male Female

Number In Thousands

Total.......................................................... 31,706 12,514 19,192

AgeUnder 15 Years......................................... 2,383 1,333 1,05015-44 Years............................................... 9,969 2,680 7,28945-64 Years............................................... 6,958 3,424 3,53465 Years and Over.................................... 12,396 5,077 7,319

RegionNortheast................................................... 7,103 2,979 4,123Midwest..................................................... 7,207 2,857 4,351South......................................................... 12,016 4,621 7,395West.......................................................... 5,380 2,057 3,323

Notes Available

Number of Discharges from Short-Stay Hospitals by Selected Characteristics (United States, 2000)

27

Selected BothCharacteristic Sexes Male Female

Number Per 1,000 Population

Total.......................................................... 114.0 92.0 135.1

AgeUnder 15 Years......................................... 39.4 43.1 35.515-44 Years............................................... 81.6 43.9 119.345-64 Years............................................... 114.2 115.8 112.665 Years and Over.................................... 359.6 352.8 364.4

RegionNortheast................................................... 135.5 117.4 152.5Midwest..................................................... 112.8 91.5 133.2South......................................................... 121.6 96.4 145.4West.......................................................... 85.4 65.4 105.3

Notes Available

Rate of Discharges from Short-Stay Hospitals by Selected Characteristics (United States, 2000)

28

National Ambulatory Medical Care Survey

Describes ambulatory care visits made to physician offices within the U.S.−

By physician practice characteristics

−

By patient age, gender, race−

By payment type and expected source of insurance for this visit

−

By patient’s principal reason for visit

29

Physician Practice Number of Visits PercentCharacteristic in thousand Distribution

All visits 823,542 100General and family practice ..................... 198,578 24.1Internal medicine ............................ 125,556 15.2Pediatrics ................................ 103,734 12.6OB/GYN…. 65,135 7.9Orthopedic surgery…. 46,155 5.6

Professional IdentityDoctor of Medicine 756,813 91.9Doctor of Osteopathy 66,729 8.1

Geographic RegionNortheast 183,029 22.2Midwest 206,727 25.1South 251,300 30.5West 182,485 22.2

Percentage Distribution of Office Visits: U.S., 2000

Notes Available

30

Population-basedDesigned to assess the health and nutritional status of adults and children in the U.S. through interviews and direct physical examinationsMedical and dental examinations, physiological measurements, laboratory tests

National Health and Nutrition Examination Survey (NHANES)

31

Medical Expenditure Panel Survey (MEPS)

Provides policy makers with up-to-date, highly detailed information on how Americans as a whole and different segments of the population use and pay for health careDocuments insurance coverage and other access issues

32

Example of Findings: MEPS, 2001

In the first half of 1999, 15.8% of all Americans were uninsuredAmong Americans under 65, 36% of Hispanics and 21% of blacks were uninsured during the first half of 1999, compared with only 14% of whitesEven though Hispanics represented only 13% of the non-elderly U.S. population, they accounted for a fourth (25%) of the entire uninsured population

33

Example of Findings: MEPS, 2001

Young adults ages 19–24 were more at risk of being uninsured than any other age group (almost a third (32%) of young adults were uninsured)During the first half of 1999, among people under age 65, those who were separated from their spouse were more likely to be uninsured (33%) than people of any other marital status

34

Medicare Current Beneficiary Survey

Interviews with about 10,000 Medicare beneficiariesNationally representative sampleData collected include:−

Self-reported health and functional status

−

Access to care−

Satisfaction with care

Survey data can be linked to claims data